The overall objective is to develop methods for the accurate a priori calculation of conformations of complicated molecules, especially steroids, by methods of classical mechanics, and to check, calibrate and improve them until they are of reliable predictive value. Once it is possible to determine accurate conformations with reasonable ease and speed, more emphasis will be placed on the calculation of observable physical properties such as optical rotatory dispersion curves and proton magnetic resonance spectra, both to improve our ability to predict these quantities, and to use such predictions for structural determinations. A number of compounds which are known to or might exhibit "anomolous" or unusual properties or types of interactions will be prepared and examined in these connections.